The Effects of High Cognitive Uncertainty on Effective Collaboration in a Post Merger Setting

In, “A distributed cognition perspective on newcomers' change processes: The management of cognitive uncertainty in two investment banks,” A. Alexandra Michel shows that in an investment banking setting, a bank which amplified cognitive uncertainty created a collective-centric organization. As she writes, “because demands exceeded individuals’ cognitive capacity, bankers used organizational resources to solve problems inductively.” Based on her conclusions, the implicit assumption is that when cognitive uncertainty is intermittent, or absent, organizations will tend to be less group-oriented in their approach to problem solving, more individualistic, and, logically, less effective at collaboration. Groups with high cognitive uncertainty problem solve inductively and will harness resources across the group more effectively.

Another implicit claim is that if cognitive uncertainty is high enough, and participants feel that solving problems is outside the realm of their own skill set, they are less likely to satisfice, and more likely to collaborate to solve problems, and therefore more likely to have better decision outcomes. Similarly, high levels of cognitive uncertainty imply low levels of taxonomic classification and, presumably, diverse member skillsets are more highly leveraged – the point of view of another member is more valued than it would be otherwise.

In other words, groups with low levels of cognitive uncertainty close off potential outcomes. As Frank Knight writes, “Uncertainty must be taken in a sense radically distinct from the familiar notion of Risk, from which it has never been properly separated…It will appear that a measurable uncertainty, or 'risk' proper, as we shall use the term, is so far different from an unmeasurable one that it is not in effect an uncertainty at all." If the future is completely unpredictable, as Knight would argue, cognitive certainty is an artifice that ignorantly presume omniscient knowledge on the part of the manager (or organization). Any attempt to reduce cognitive uncertainty, in Knightean terms, implies the ability to predict the future. The optimal solution may be unintentionally closed off, or muted by the restriction of options.

Our question in this installment of the post merger integration blog is what about the case in which two organizations have merged together? This case adds some complexities that Michel’s does not possess. Knight would argue that it doesn’t matter. Any attempt by a leader to control the group norms would result in the reduction of potential value of the group. In effect, through directing, a manager would reduce organizing, and potentially limit or interfere with sensemaking.

One hypothetical situation is that when two cultures come together, the groups may come into a new situation with bounded senses of cognitive uncertainty to a certain degree. While they may not know what the new group norms are, they know what the norms from their legacy institutions were. Ironically, in this case, high levels of cognitive uncertainty from a legacy organization could become a norm. If I expect high levels of uncertainty in the newly formed organization, I am clinging to a norm, and thereby my cognitive uncertainty which was high in the old organization is lowered in the new one. Another practical complication is that without some cognitive certainties, a group formed by two organizations may never integrate. That is to say, if it isn’t forced on the members, they may not ever feel the need to do it.

This could imply that in a post merger setting some of what researchers argued that organizations should reduce cognitive uncertainty to compensate for bounded rationality (March and Simon, 1958) may apply. In other words, unless I tell the groups they have to integrate they may never do it (at the worst) and potentially will waste time while integrating (at the best). The research we have seen in ingroup/outgroup psychology could support this hypothesis through the fact that a merger creates groups that may have not been present pre merger through the creation of “legacy x” and “legaxy y” identities. For example, when two companies that operated with high degrees of uncertainty pre-merger (and presumably have low levels of “within group similarity of attitudes, understanding, and language” (Weick and Roberts, 1993: 358) the legacy orientation creates group identities that may need to be broken down externally by managers to re-establish high levels of cognitive uncertainty.

We are interested in testing Michel’s thesis in a post merger environment. Our experiment will be to take two highly functioning teams prescreened as operating with high levels of cognitive uncertainty. (In some ways this could be simply a newly formed group). The premise of the experiment would be to have the groups complete a game separately and then split the groups in half and create merged groups of equal representation in terms of numbers and ability from each legacy group to repeat the same game. We would tell one group that they have to collaborate with members from both legacy groups as well as tell them who from the previous round performed best on the task and which team scored higher than the other. We would tell the other merged group nothing and observe both groups as they complete the same game for the second time. (A repeated game would simulate Michel’s pitch process in that it is a highly repeated activity in which performance can be improved upon with collaboration. Disclosing scores would simulate the “star culture” she outlines in the case of Red Bank).

Our hypothesis is that the unprimed (i.e. group with the highest degree of cognitive uncertainty) will perform better on the repeated game than that which was given some reductions in cognitive uncertainty. Our theory is that best practices and individual performances which were present pre-merger will still be present post-merger, but that highlighting them in one group will stifle lessons learned from the previous iteration and the emergence of new talents which were not present in the first iteration. This structure could be repeated with numerous groups to generate statistically relevant data, using scores on the game as a proxy for performance. Or we could approach ethnographically, and observe a smaller number of groups, thereby yielding comparative data to that of Michel’s. Both methods have merit in their ability to test this theory.